The envelope of gram positive bacteria consists of the following components (from inside to outside):

a) cytoplasmic membrane (proteins & phospholipids)

b) cell wall (peptidogycan & teichoic acid)

c) capsule (usually polysaccharides)

Gram-negative bacteria contain the following layers:

a) cytoplasmic membrane (proteins & phospholipids)

b) cell wall (peptidoglycan)

c) periplasmic space (contains MDO, transport proteins & degradative enzymes)

d) outer membrane (lipopolysaccharide, proteins, & phospholipids)

e) capsule (usually polysaccharide)

 

The Cytoplasmic Membrane

This is a typical biological membrane and consists of a phospholipid bilayer with proteins attached. Extrinsic membrane proteins are attached to the surface of the membrane whereas intrinsic membrane proteins are inserted into (or completely through) the lipid bilayer. Three main classes:

Cytoplasmic Membrane Proteins Used In Energy Transduction

Aerobic Dehydrogenases & Oxidases

NADH dehydrogenase, Succinate dehydrogenase, Glycerol3P dehydrogenase, Lactate oxidases ("dehydrogenases"), Pyruvate oxidase, Proline oxidase

Anaerobic Dehydrogenases

Formate dehydrogenase, Glycerol3P dehydrogenase, Hydrogenase

Terminal Reductases

Cytochrome o complex, Cytochrome d complex, Nitrate reductase, Fumarate reductase, Trimethylamine oxide reductase

Enzymes Using Proton Motive Force

ATP synthase, Flagellar motor, Energy linked transhydrogenase, Many permeases

Glycolipids are found in eukaryotes and some bacteria, sulfolipids are mostly found in photosynthetic membranes. It is the phospholipids which comprise the lipid portion of the cytoplasmic membrane of most cells, including bacteria. The outer membrane of gram-negative bacteria contains phospholipid in its inner half and lipopolysaccharide in its outer half. The lipid portion of the lipopolysaccharide is called lipid A and is an unusual type of glycolipid.

Phospholipids of Bacteria Glycerol + 2 fatty acids + phosphate + base.

The fatty acids of E. coli are mostly palmitic (16:0), palmitoleic (16:1), and cis-vaccenic (18:1). These all contain an even number of C atoms. Traces of longer and shorter fatty acids also occur.

Palmitic acid (16:0) CH3 (CH2)14 COOH

Palmitoleic acid (16:1) CH3 (CH2)5 CH=CH(CH2)7 COOH

cis-Vaccenic acid (18:1) CH3 (CH2)5 CH=CH(CH2)9 COOH

Oleic acid (18:1) CH3 (CH2)7 CH=CH(CH2)7 COOH

Naturally occurring unsaturated fatty acids are of the cis configuration. Eukaryotes usually have oleic acid instead of cis-vaccenate. Oleate and cis-vaccenate differ only in the position of the double bond. Eukaryotes and cyanobacteria have fatty acids with 2 or more double bonds, whereas most bacteria contain only mono-unsaturates.

Synthesis of Fatty Acids. (see Diagram)

a) Intermediates are attached to acyl carrier protein (ACP) via a prosthetic group known as phosphopantetheine, which also forms part of Coenzyme A. ACP has 77 amino acids, i.e. a very small protein. It carries intermediates in FA synthesis whereas CoA carries intermediates in FA degradation.

Coenzyme A: HS-Phosphopantetheine-P-P-Adenosine-3'P

ACP: HS-Phosphopantetheine-P-O-Serine-ACP

H H OH CH2 O
fatty acid | | | || ||
here ÆÆ HS-CH2CH2-N-C-CH2CH2-N-C-CH-C-CH2-O-P-OH ¨¨ to rest of CoA/ACP
|| || | |
O O CH3 O

b) Precursor for FA synthesis is acetyl CoA. Chain initiation: acetyl group transferred to ACP (reaction 2 on diagram) from acetyl CoA. Chain elongation: acetyl CoA converted to malonyl CoA by incorporation of CO2 (reaction 1). Malonyl group transferred to ACP (reaction 3). Condensation of malonyl group with growing FA chain (reaction 4) is driven by release of CO2. Reduction of ketone to CH2 (steps 5,7, and 8) produces an acyl-ACP. This is then elongated, two C's at a time by further operation of cycle.

c) Insertion of double bond may occur by two mechanisms. Anaerobes and facultative anaerobes, simply omit a reductive step during synthesis. When precursor reaches C10, they produce a cis double bond (step 6) which cannot be further reduced (by step 8 which can only reduce trans double bonds). Further elongation gives palmitoleic and cis-vaccenic acids. The step 6 enzyme can isomerize cis/trans at the double bond thus producing a mixture of cis and trans C10 precursors. The trans precursors go on to become saturated FA, while the cis ones become unsaturated. Note that step 7 has two enzymes: a "short" isoenzyme for C8 and below and a "long" isoenzyme for C12 and above.

H
|
-C==C- ÆÆ SFA
|
OH H H
| |
-C----C- ÆÆÆ -C==C- ÆÆ UFA
| | | |
H H H H

d) Obligate aerobes introduce the double bond after full elongation of the FA chain. Hydroxylation using molecular O2 is followed by removal of H2O to leave C==C. May be repeated to give polyunsaturates. Similar O2 dependent hydroxylation is required for steroid synthesis.

H H O2 HO H H2O
| | \ | | /
C---C ÆÆÆ C---C ÆÆÆ C==C
/ \
NADPH2 NADP

e) Cyclopropane FA: In stationary phase E. coli converts its unsaturated FA to cyclopropane FA. The methylene group comes from the one carbon pool via S-adenosyl methionine. Thought to be a protective mechanism but this theory is unproven.

CH2
C-1 unit / \
-HC==CH- ÆÆÆ -HC---CH-

Phospholipid head groups.

There is considerable choice in the organic "base" linked to the phosphate. In E. coli approximately 70% phosphatidyl ethanolamine (PE), 20% phosphatidyl glycerol (PG) and 10% cardiolipin. PG is converted to cardiolipin especially in the stationary phase. Exponential cells have perhaps 5% cardiolipin; fully stationary cells have up to 25%. Other organisms may possess phophatidyl choline (major phospholipid in animals), phosphatidyl inositol or other phospholipids.

Phospholipid Synthesis: (see diagram)

(It was originally thought PG combined with CDP-diglyceride to form cardiolipin, as shown in diagrams in some books. Confusion is due to fact that PG donates pieces to cardiolipin, lipoprotein, and membrane derived oligosaccharides -- only recently appreciated -- see general lipid metabolism diagram.)

Temperature. Lipid bilayer must be in fluid state for membrane proteins to function. Fluidity refers to the state of the fatty acyl chains. Membranes may be viewed as viscous two-dimensional liquids in which proteins float. Incorporation of eg lactose permease will occur into a rigid lipid bilayer, but the protein is then non-functional. Fluidity increases with content of unsaturated fatty acids (UFA), and with temperature. The mixture of saturated and unsaturated acids keeps the membranes fluid over the whole temperature range for growth. The ratio of UFA to SFA increases as temperature falls.

In E. coli the change is due to increase in synthesis of only cis-vaccenic acid (rather than both of the UFA). E. coli contains two fatty acid condensing enzymes, which catalyze the elongation reaction. Enzyme I works well up to the C14 precursor for C16 products. It cannot elongate C16 to C18. Enzyme II can elongate C16 to C18, but only if C16 precursor is unsaturated. Hence, all the cis-vaccenic acid produced in E. coli is due to operation of condensing enzyme II. Enzyme II is inherently sensitive to temperature. At low temperatures, a lot of cis-vaccenate is made; at high temperatures, much less.

In Bacillus, where double bonds are made by the O2 insertion pathway the desaturation enzymes are induced by a drop in temperature and inactivated by an upshift.

PHOSPHOLIPID SYNTHESIS IN E. COLI

THE PEPTIDOGLYCAN

The cell wall of gram-negatives contains one giant molecule of peptidoglycan (or mucopeptide) also called the murein sacculus (since it is bag-shaped). In gram-positives there are multiple layers of peptidoglycan linked to teichoic acids. This is the rigid layer of the cell envelope.

Synthesis of peptidoglycan illustrates two important general principles in polysaccharide metabolism:

Stages (1) and (2) actually occur together:

a) Fructose-6-P + Glutamine = Glutamate + Glucosamine-6-P

b) Glucosamine-6-P + Acetyl CoA = NAG-6-P

c) NAG-6-P = NAG-1-P

d) NAG-1-P + UTP = UDP-NAG + PPi

e) UDP-NAG + PEP = Pi + UDP-NAG-enol pyruvyl ether

f) UDP-NAG-enol pyruvyl ether + NADPH2 = NADP + UDP-NAM

Stage (3): Pentapeptide

Add amino acids to carboxyl group of the lactyl side chain of NAM. Each addition requires a unique enzyme and ATP. L-Ala, D-Glu, mesoDAP, D-Ala-D-Ala. Note: DAP is connected via its alpha-NH2 to the gamma-COOH of glutamate. The alpha-COOH of this D-Glu is amidated at a later stage. Final addition is of a pre-made dipeptide of two D-Alanines. The amino acid sequence of the pentapeptide given is for E. coli. The antibiotic D-cycloserine inhibits both steps of D-Ala metabolism:

a) Racemase: L-Alanine = D-Alanine

b) Synthetase: D-Ala + D-Ala = D-Ala-D-Ala

Stages (4) and (5): Bactoprenol Carrier

Previous reactions occur in cytoplasm. Bactoprenol phosphate is a membrane lipid which carries precursors across cytoplasmic membrane and allows assembly of polysaccharide chain on outer surface of cytoplasmic membrane. Bactoprenol is a polyisoprenoid. Sometimes called undecaprenol since has 11 isoprene units of 5 carbons each. In eukaryotes the carrier lipid is dolichol-phosphate with twenty isoprene subunits.

H--[CH2-C=CH-CH2]11--O-PO3(2-)
| Bactoprenol Phosphate
CH3

In Staphylococcus (and many cocci) pentaglycine bridge added here. Uses tRNA but not ribosomes. Glycines are added to the side chain NH2 of lysine. Note that assembly is reversed relative to direction in protein synthesis. In E. coli (and most gram negative rods and also the gram positive bacilli) the lysine is replaced by meso-2, 6-diaminopimelic acid and there is no pentaglycine bridge.

CH2NH2 HOOC-CH-NH2
| |
(CH2)3 (CH2)3
| |
HOOC-CH-NH2 HOOC-CH-NH2
Lysine Diaminopimelic acid (DAP)

NAG/NAM-pentapeptide is transferred to the growing polysaccharide chain. Formation of beta-linkage from 1-position of NAM to 4-OH group of last NAG residue in chain. Bactoprenol-pyrophosphate is released and is converted back to the single phosphate form by a specific pyrophosphatase.

7) Cross Linking

Cross-linking locks peptidoglycan chains into final position in cell. The cross-linking reaction occurs outside the cytoplasmic membrane where there is no supply of ATP. Energy comes from hydrolysis of terminal D-Ala-D-Ala link. The outermost D-Ala is lost and the inner D-Ala is linked either to the fifth glycine (Staph) or to the side chain carboxyl of DAP (E. coli). In Staph multiple random cross links between many chains. In E. coli relatively few cross links and many side chains are not cross-linked. Instead DD-Carboxypeptidase I removes terminal D-Alanine so "inactivating" some side chains.

Antibiotics Which Affect Cell Wall Synthesis

Phosphonomycin PEP analog, inhibits condensation of PEP with UDP-NAG.

D-Cycloserine Analog of D-Alanine. Inhibits both D-Ala racemase and D-Ala-D-Ala synthetase. Binds about 100 fold better to these enzymes than D-Ala itself. Rigid structure holds it in one particular conformation-presumably the same conformation that D-Ala takes up when in enzyme active site.

Bacitracin binds polyisoprenyl pyrophosphates, including bactoprenol pyrophosphateand prevents them from being converted back to the single phosphate form required as active carrier. Bacitracin also stops synthesis of other polysaccharides which require carrier lipids. Needs Mg2+ to bind bactoprenol-PP. Outer membrane of gram-negatives protects against bacitracin.

Beta-Lactams (Penicillins and Cephalosporins) Contain a beta-lactam ring fused to another ring (5-membered in penicillins, 6-membered in cephalosporins). Original penicillin contained a mixture of side chains on the penicillin nucleus. If phenylacetic acid is added to fermenting mold, most of product is benzyl-penicillin (Pen G). Need 1000x as much to kill gram negatives as gram positives, - due to exclusion by outer membrane.

Cephalosporin C - from different mold. If the penicillin precursor has its sulfur containing ring formed by a different mechanism to give a six membered ring this yields cephalosporins. Cephalosporin C itself is of no practical use, however, many derivatives are used.

b-Lactams prevent cross-linking of peptidoglycan. They are transition state analogs of D-Ala-D-Ala. The labile bond (C-N) of the beta-lactams corresponds to the C-N bond broken during the transpeptidation reaction in which the terminal D-Ala-D-Ala is used to generate cross-links.

Penicillin treatment reduces degree of crosslinking. Peptidoglycan layer falls apart. Beta-lactams only affect cells which are actively growing (i.e. making and crosslinking peptidoglycan). Pre-made peptidoglycan is not affected. Cells are only killed if the external medium is hypo-osmotic. Beta-lactam treated cells may be saved from lysis by high salt or sugar concentrations.

Enzymes recognizing D-Ala-D-Ala can bind b-lactam instead. They then break the lactam (C-N) bond and form a covalent enzyme-substrate intermediate with the beta-lactam molecule. In the case of the real substrate the intermediate reacts further. In the case of penicillin the reaction product is stable, i.e the enzyme is permanently inactivated.

Some beta-lactams

Penicillin binding proteins (PBP) of E. coli:

PBP# MW (kd) Enzyme Activity Function

1 91 Transpeptidase Elongation

2 66 ? Cell Shape

3 60 Transpeptidase Cross-Wall

4 49 Carboxypeptidase II Attaches Lipoprotein

5 42 DD-Carboxypeptidase Removes last D-Ala

6 40 DD-Carboxypeptidase Removes last D-Ala

Most of original b-lactams are most active against PBP-1 and cause lysis. Mecillinam binds best to PBP-2 and result is formation of giant spherical cells, followed by lysis only at higher antibiotic concentrations. Cephalexin favors both PBP-2 and PBP-3 and produces filaments with bulges.

Lysozyme An enzyme not an antibiotic. Found in tears, sweat, egg white, etc. Hydrolyses the sugar chains of peptidoglycan between NAG and NAM.

PERIPLASMIC SPACE

Between peptidoglycan and outer membrane. The periplasmic space is 20 to 40% of total volume of cells grown under typical conditions. It contains the following:

OUTER MEMBRANE

ECA is an acidic polysaccharide of NAG, N-acetyl-mannosamine uronic acid and 4-acetamido-4,6-dideoxy-D-galactose. ECA is covalently linked to a phospholipid molecule in the OM. Colanic acid contains glucose, galactose, fucose and glucuronic acid. It is made in response to adverse environmental conditions (high OP, low temperature, or dessication). ECA and colanic acid are made by most enterobacteria. K antigens are specific capsular polysaccharides made by most E. coli but not bySalmonella. Common lab strains lack K-antigens. (Do not confuse K-antigens and E. coli K-12, the two K's have nothing to do with each other). Some K-antigens are anchored to the OM by lipid tails.

Lipopolysaccharide (LPS). Three regions:

Synthesis of LPS: Synthetic enzymes all located in cytoplasmic membrane. Lipid A synthesized first and core added by sequential addition of sugars from NDP-Sugar donors. The O-antigen is assembled on the same bactoprenol carrier lipid as used for peptidoglycan synthesis.

Lipid A and the KDO region of LPS are essential for viability. The O-antigen and rest of core down to HEP are not. If core of LPS is damaged OM becomes permeable to many substances otherwise excluded. If O-antigen absent cell much more susceptible to animal immune system. Large variety of O-antigen structures (over 150 in E. coli ). Varying O side chain allows bacteria to stay ahead of immune recognition. Assembly occurs on cytoplasmic membrane. LPS is translocated to OM via the IM/OM adhesion sites - the Bayer patches.

Bayer sites. Temporary discontinuities in peptidoglycan allow limited regions of adhesion between OM and IM. These are sites of entry for phage DNA in many cases. Structure of Bayer site uncertain - also unknown whether they are short lived or permanent. Observed under EM by phage attachment and by binding of antibody/ferritin complex to newly made O-antigen.

Structure of Lipid-A: beta-1,6-Diglucosamine with amino groups substituted by the 14-carbon hydroxy-fatty acid, beta-hydroxymyristate (BHM). The -OH group of BHM is substituted by further fatty acids (saturated C12 and C14, traces of C16 and even some BHM). Ethanolamine and 4-aminoarabinose are sometimes attached. Divalent ions and polyamines are tightly bound and form ionic bridges between phosphate groups of LPS molecules in lipid A and inner core regions. Lipid A is responsible for the endotoxin activity of LPS versus eukaryotic cells.

Phage Receptors Both LPS and OM proteins may act as receptors for bacteriophages. Receptors for P1, T4, and T7 include both protein and LPS components. Mutants defective in LPS sugars change their phage susceptibility. Thus E. coli O8 adsorbs phage Omega8 (binds to the O8 O-antigen) whereas E. coli K-12 (no O-antigen) does not. E. coli K-12 phages specific for the LPS include:

U3 requires presence of GAL on end of core

C21 only adsorbs if final GAL absent exposing GLC below

Br10 requires phosphate on heptose

Br2 requires inner GLC(GAL)-GLC of core

FP1 & T4 require heptose of inner core.

Separation of cytoplasmic and outer membrane (Method of Osborn):

Outer Membrane Proteins

a) Lipoprotein. Approximately 700,000 molecules per cell, i.e. the most abundant protein (numerically) in E. coli. Part is covalently attached to peptidoglycan. Holds cell wall and OM together. MW7000. 58 amino acids. Linked by -NH2 group of C-terminal Lys to carboxyl group of every 10-12th (on average) DAP residue of peptidoglycan. N-terminus is glyceryl-Cys to which two fatty acids are attached by ester links (to the glycerol) and one fatty acid by amide link (to the cysteine). For every bound lipoprotein there are two unlinked to the peptidoglycan.

b) OmpA Protein. Structural protein. Receptor for the F-pilus during mating. Present in large amounts.

c) Porins. Form trimeric pores through OM. Approx 200,000 per cell. The pores exclude molecules of more than 700 in MW. E. coli K-12 has two major porins; OmpF (38Kd) decreases in high OP whereas OmpC (37Kd) increases in high OP. Salmonella typhimurium has a third porin OmpD (34Kd).

Mutants lacking OmpC and OmpF proteins grow poorly unless growth substrate is present in high concentration. They are more resistant to e.g. Cu2+ or cephaloridine which can no longer enter so easily.

Under certain circumstances other porins may be expressed e.g. during phosphate limitation the PhoE porin which favors the entry of phosphate or similar anions is produced in large amounts. Maltose induces the LamB porin. Porins cover approximately 60% of cell surface and LPS approximately 30%.

Porin MW (Kd) Function Receptor for

OmpF 38 major porin TuIa, T2, ColA

OmpC 37 major porin TuIb, Mel, 434

PhoE 37 phosphate porin TC45

LamB 47 maltose porin Lambda

d) Minor Proteins. Mostly receptors for molecules too large to get through pores formed by porins. Most are also receptors for phage and colicins as are the major proteins (except lipoprotein). A few degradative enzymes, e.g. phospholipase are the only enzymes in the OM.

Protein MW (kd) Transport of Receptor used by

Tsx 26 nucleosides T6, ColK

BtuB (Bfe) 66 Vit B12 BF23, ColE

Cir 74 ferric iron ColI, ColV

Iut 75 Fe-aerobactin CloacinDF13

(on ColV plasmid)

FhuE 76 Fe-coprogen

FhuA (TonA) 78 Fe-ferrichrome Tl, T5, f80, ColM, albomycin

FecA 81 Fe-citrate

FepA 81 Fe-enterobactin ColB, ColD

Fiu 83 ferric iron

______________________________________________________

Vitamin B12 receptor (Bfe protein). Usually about 200 per cell. Phage BF23 only needs one or two to kill cells. E. coli has only two coenzyme B12 using enzymes, neither of which is essential. 1) Alternative synthetic route to methionine involving transfer of CH3 from tetrahydrofolate to homocysteine. 2) Growth on ethanolamine as N-source requires ethanolamine ammonia lyase. E. coli cannot make B12 and since MW is 1356 it cannot enter by porins. Receptor is a glycoprotein (contains Glc, Gal, Rha, uronic acids and amino sugars). Uptake process:

a) Fast initial binding to receptor (no energy needed)

b) Slower internalization of bound B12. Energy dependent, inhibited by energy poisons such as cyanide and uncouplers, but not by ATP deficiency, i.e. this is a direct PMF driven system. Energy coupling is due to the TonB protein.

c) The cytoplasmic membrane receptor for B12 is the BtuC protein (discovered after the diagram was produced!).

The TonB protein energizes uptake of B12 by Bfe and of Fe by the enterochelin and ferrichrome systems and is also necessary for infection by some phages which use these receptor systems (see diagram). All these systems have OM receptors and cytoplasmic membrane permeases and operate like the B12 system with respect to energy requirements. The Bayer sites appear to be involved. Note: phage-BF23 and ColE don't need TonB although they use Bfe.

Iron Transport

Enterochelin and ferrichrome are both Fe chelators and each has a separate receptor system energized by TonB. There is also a citrate inducible Fe transport system in which Fe is taken up chelated by citrate. There are also several other proteins of uncertain function which are affected by presence of Fe. Biological iron chelators, known as siderophores, are of two types: catechols and hydroxamates.

Enterochelin (= enterobactin, a catechol) is a cyclic trimer of 2,3-dihydroxybenzoylserine (DHBS), OM receptor is FepA protein. The FepB protein is the inner membrane permease. The Fes protein is an esterase which hydrolyses Fe3+ - enterochelin into DHBS-Fe3+ monomers. Such monomers may be easily reduced to the Fe2+ form which is used in biosynthesis. E. coli synthesizes enterochelin.

Ferrichrome (a hydroxamate) is not synthesized by E. coli but only by fungi, e.g. Penicillium. Despite this, many bacteria have receptors for Fe3+/ferrichrome, hence stealing the fungal siderophores. Albomycin is an antibiotic analog of ferrichrome, made by the fungi to kill such dishonest bacteria! OM receptor is FhuA (TonA), inner membrane receptor is FhuB.

Citrate allows uptake of Fe if high concentration of Fe is present. System is induced by citrate plus Fe. Salmonella cannot use citrate for Fe uptake but uses it as a C-source. E. coli does the reverse. System requires TonB.

Aerobactin (a hydroxamate) and its receptor system.are made by E. coli carrying the ColV plasmid (also make colicin V). Such E. coli are often virulent - due, in part, to aerobactin which can scavenge Fe from transferrin, the iron binding protein of blood.

Biological Fe Warfare: Human serum contains antibody to enterochelin. Certain E. coli can grab Fe from transferrin and other human Fe proteins using aerobactin (most other bacterial siderophores bind iron less strongly than transferrin). They also often make ColicinV which eliminates competing E. coli by binding to the Cir receptor (Cir is an uncharacterized Fe uptake system). Enterobacter cloacae carries Cloacin plasmid CloDF13. Cloacin DF13 kills E. coli which are using the aerobactin receptor. E. cloacae uses aerobactin itself but makes a cloacin immunity protein to protect itself. Most enterobacteria steal ferrichrome made by Penicillium which in revenge makes albomycin which enters by the ferrichrome receptor system and is a lot more effective against gram negatives than penicillin.

EXPORT AND PROCESSING OF ENVELOPE PROTEINS

Proteins of the OM, the periplasmic space and the IM must be transported into the IM. Those destined for the OM or periplasmic space must then be transported further out into the envelope.

Signal hypothesis (by Blobel):

Example: maltose binding protein (periplasmic, malE product). Precursor is 26 amino acids (i.e. the signal sequence) longer than mature protein. If a hydrophobic amino acid in the middle of the signal sequence is changed, by mutation, to a charged residue, the malE precursor is NOT exported and becomes a cytoplasmic protein. Conversely, if a gene fusion is constructed in which the N-terminus of malE is joined to the C-terminal region of lacZ (the gene for b-galactosidase which is cytoplasmic), then a hybrid protein, which has the malE signal sequence and the lacZ enzyme activity is found in the membrane. Note that the malE-lacZ fusion protein is NOT periplasmic, i.e. export goes as far as the IM but cleavage of the signal sequence and release into the PP-space does not succeed. Similar experiments have been done with lamB (maltose porin, an OM protein) and malF, (IM maltose permease). In these cases, the hybrid protein ends up in the OM or IM respectively. Thus a signal sequence conveys a protein INTO the membrane. For PP-space protein to get through the membrane and be released on the other side requires a protein sequence which allows complete transit - this is found in PP-space proteins but NOT in cytoplasmic or IM proteins. The characteristics which determine total export or continued residence in the membrane are NOT cleaved off but are an integral part of the mature protein.

Signal hypothesis and cotranslational export applies to most periplasmic proteins e.g alkaline phosphatase, maltose and arabinose binding proteins, several amino acid binding proteins, beta-lactamases and several excreted proteins in gram positive organisms eg diphtheria toxin and penicillinase from Bacillus. Also OM receptors, lipoprotein, ompA and ompF proteins. In eukaryotes, cotranslational export occurs across the membranes of the endoplasmic reticulum. When gene for preproinsulin is put into E. coli correct export across the IM occurs and cleavage of preproinsulin to proinsulin by the E. coli leader peptidase happens at the correct position. E. coli can also secrete ovalbumin correctly. Conversely yeast cells correctly process prokaryotic beta-lactamase. Thus the export process is highly conserved between diverse organisms. For IM proteins, the situation is more complicated. Some (e.g. malF permease) follow the above procedure. So do the penicillin binding proteins, PB5 and PB6. Many others do not follow the signal/cotranslation procedure.

a) Bacteriophage M13 coat protein accumulates in the IM of E. coli before assembly into virus particles. Pre-coat protein is made without membrane insertion. It then goes to IM where cleavage of signal sequence and membrane insertion occur post-translationally. Although a lot of work has been done on M13 coat-protein and much has been made of it, this protein is probably a unique anomaly - its ultimate position is in a virus particle, not the cell envelope.

b) Many IM proteins have no signal sequence, eg. lactose permease, the integral membrane proteins of the ATPase, and the histidine transport system. These probably fold into a structure which naturally partitions into the hydrophobic membrane. Note that none of these proteins ever lose contact with the cytoplasm - they are not really exported.

Signal Peptides. About 20 have been sequenced. Little homology. Positively charged, basic N-terminus of 2 to 8 amino acids, then a long stretch of hydrophobic amino acids. The amino acid just before cleavage site has a short side chain. No similarity in structure beyond cleavage site. Exist as alpha-helix structures. Mutational insertion of charged AA or proline (helix breaker) into hydrophobic stretch destroys signal function.

Rest of Protein.

a) Long sequences of hydrophilic amino acids may prevent complete transit across membrane even if protein is provided with a signal sequence (e.g. malE-lacZ hybrid).

b) C-terminal region of PP-space protein may be necessary to allow protein to be released from IM after transit, i.e. protein must be able to dissolve in the PP-space after cleavage of signal sequence. This was shown with b-lactamase where C-terminal region mutations allow crossing of IM but prevent release of soluble enzyme into PP-space. (Originally, this was used as evidence against signal hypothesis as no was protein released. However, beta-lactamase was detected attached to the outer side of IM).

Export Machinery (See diagram)

ISOPRENOIDS

Successive condensations of the 5-carbon isoprene unit give molecules with multiples of this branched C5 repeating unit which usually contain one double bond per C5.

A wide variety of products are produced:

Molecular fossils.